【摘要】 现有方法在舰船物联网威胁识别率与缺陷识别率上表现不佳,因此提出一种改进卷积神经网络的舰船物联网安全风险估计方法。对舰船物联网安全风险数据进行分类也就是分类安全风险指标。对于动态类安全风险数据,需要对其进行数据补充。运行类安全风险数据的预处理需要进行数据清理。基础类安全风险数据的预处理需要进行数据变换与数据归一化处理。基于改进卷积神经网络提取舰船物联网安全风险数据特征,使用的改进卷积神经网络为SSD神经网络。基于灰色层次分析、Borda序列、风险矩阵构建舰船物联网安全风险估计模型。对设计方法进行实践应用,测试其舰船物联网威胁识别率与缺陷识别率,结果表明该方法取得了识别率数据上的突破,能够保障舰船物联网的安全。更多还原

【Abstract】 The existing methods do not perform well in the threat recognition rate and defect recognition rate of the ship IoT. Therefore, an improved convolutional neural network security risk estimation method for the ship IoT is proposed. To classify the security risk data of ship IoT is to classify security risk indicators. For dynamic security risk data, it needs to be supplemented with data. The preprocessing of operational safety risk data requires data cleaning. The preprocessing of basic security risk data requires data transformation and data normalization. Based on the improved convolutional neural network to extract the security risk data characteristics of the ship Internet of things, the improved convolutional neural network used is the SSD neural network. Based on the gray analytic hierarchy process, Borda sequence, and risk matrix, the ship IoT security risk estimation model is constructed. The design method was applied in practice to test the threat recognition rate and defect recognition rate of the ship’s Internet of Things. The results show that the method has achieved a breakthrough in the recognition rate data and can guarantee the safety of the ship’s Internet of Things.更多还原